Nucleotide bias in SS/DS RNA viruses vs. retroviruses

[Ulrich Melcher]

Several points regarding D. Shivak's recent query (copy appended below):

1.  Kypr and Mrazek (Nature 327:20, 1987) do mention a high A content for 
influenza, human papilloma, papova RNA viruses).

2.  They also mention a high A content for cauliflower mosaic virus.  Kelly 
Chenault and I have also noted that, but later (Biochimie 76,3-8, 1994).  
Indeed, the near-uniform high A content is present in all sequenced 
caulimoviruses and badnaviruses.  These are retroid viruses of plants.  The 
content is very similar to that seen in the lentiviruses.

3.  Our analysis of CaMV nucleotide sequences suggested that the main 
contributing factor to the high A content is a suppression of A to G 
transitions.  These were significantly lower than G to A or C to T and T to C 
transitions.  Of course, we have no clue as to how that suppression occurs 
biochemically.

4.  The nucleotide composition bias in viruses does, as Bronson and Anderson 
point out, influence the outcome of phylogenetic analyses.  Chenault and I 
noticed that in a phylogenetic study of CaMV isolates ( J. Mol. Evol. 39: 496, 
1994).

5.  The idea proposed by Bronson and Anderson that the composition biases arose 
because of competition between two variants on their way to speciating can be 
questioned when the plant viruses are included in the thinking.  The caulimo and
the badnaviruses are the only two known types of retroid viruses in plants.  The
proposed mechanism would require that they have opposite base compositions.  
They do not.  Alternatively, one could argue that the plant retroid viruses 
evolved from a lentivirus after lentiviruses diverged from oncoviruses.  One 
could argue that, but I don't think many would.  Anyone else have any ideas on 
why lentiviruses, caulimoviruses and badnaviruses should share high A contents?

6.  My interpretation of the Bronson & Anderson article differs a little from 
Shivak's  "mutational bias due to reverse transcriptase in retroviruses may play
an important role in speciation".  I think B & A were arguing that competition 
was driving speciation.  The competition allowed two variants with complementary
base compositions to co-exist.  As a consequence of the variant compositions the
RT adjusted its specificity.  True, this is a subtle difference.  Yet, I think 
it important.

In message <D02J1u.H1x at sue.cc.uregina.ca> Dave Shivak writes:
> Network was doing some odd things last night when I posted this, so 
> apologies if this is a repeat posting....
>  
> I was wondering if any recent work has been done on SS or DS RNA virus 
> phylogeny specifically mentioning nucleotide composition and 
> nucleotide bias of RNAdepRNApol as an evolutionary influence.  I 
> recently read the paper _Nucleotide composition as a driving force in the 
> evolution of retroviruses_ (Bronson & Anderson (1994) J. Mol. Evol. 
> 38:506-532 article) which suggests mutational bias due to 
> reverse transcriptase in retroviruses may play an important role 
> in speciation.  I was curious as to whether these results may be 
> generalized to other viral systems with similar low-fidelity nucleic 
> acid replication, such as the SS and DS RNA viruses.  Any references 
> or comments of use would be much appreciated (please Email). 
>  
> Thanks in advance
> Dave Shivak
> shivakd at meena.cc.uregina.ca
>  
> 


Ulrich Melcher                   biocukm at okway.okstate.edu
Department of Biochemistry       umelcher at bmb-fs1.biochem.okstate.edu
    and Molecular Biology        tel. 405-744-6210
Oklahoma State University        FAX  405-744-7799
Stillwater OK 74078-0454